Why do we age, and why do some individuals age faster than others? Genetic studies have found many genes that can extend lifespan in laboratory populations. However, these genes explain little of the substantial genetic variation in lifespan that we observe in natural populations, including humans. Environmental studies have found that diet restriction (DR) can extend lifespan, and that this effect is highly conserved across taxa. However, within populations, the DR response shows considerable genetic variation. As with lifespan, the pathways that account for this genetic variation in the DR response are also unknown. Our overarching goal is to understand the genetic pathways, functional mechanisms and selective forces that shape this natural variation in aging. Towards this end, we have gathered three important sets of preliminary data. First, we have found that sex, diet, tissue type, genotype, and age all have substantial effects on the fruit fly metabolome. Second, diet restriction, which can extend mean lifespan, leads to a dramatic reversal of the effect of age on the metabolome, and this reversal appears to be associated with glycogen, glucose, and tryptophan metabolism. Third, diet restriction extends lifespan in some genotypes, while in others there is no response at all. Based on these data, we hypothesize that genetic and environmental factors that extend lifespan do so predominantly by slowing age-related changes in metabolic pathways, and that variation in these pathways will allow us to a) predict the longevity of a genotype; b) predict whether lifespan in a given genotype will respond to DR; and c) discover the mechanisms through which DR extends lifespan. Specifically, we hypothesize that DR will extend lifespan by slowing age-related changes in the same molecular pathways that account for natural variation in longevity. We will test our hypotheses by genetically mapping the metabolome in a population that shows significant variation in rates of aging, and by identifying the causal metabolic pathways that determine how individual genotypes respond to DR. Finally, we will take advantage of the power of fly genetics to manipulate metabolite levels and gene expression in flies. These manipulations will allow us to test specific mechanistic hypotheses that arise from our preliminary studies and from results generated by the work proposed here. This innovative approach combines highly sensitive metabolomic profiling with both quantitative and molecular genetics in an easily manipulated model organism, allowing us to understand the molecular mechanisms that underlie natural variation for aging and aging-related perturbations at an unprecedented scale and level of detail. This work is expected to provide critical insights into the functional mechanisms by which well-studied factors increase lifespan, and to lead to a clearer understanding of how variation in fitness traits is generated and maintained in natural populations. The metabolomic profiling proposed here can also be carried out easily in human populations, and as such, our approach has the long- term potential to reveal the molecular pathways associated with aging and age-related disease in humans.